Help me to design a branchline coupler

hi,
i want to design a branchline coupler. the two outputs S31 and S41 have a phase diffence of 90 degrees. is it possible to put this information in the OPT block as this function: PS31-PS41=90°. my branchline coupler has to work in a frequency range of 1 to 2.5 GHz.
Hope to find here some help.

Thanks

Re: branchline coupler

Hi,

depending on a simulator you can input this optimization criteria. However, there is a smarter way to optimize branch line and rat race couplers. You can use the result from the theory that any 4 port circuit will act as a hybrid coupler if all the ports are perfectly match at a central frequency of operating band (1.75 GHz in your case). So, you will be more likely to get good phase difference if you set all your lines equal in length and only specify S11 to S44 to be -40 or less as a optimization criteria. Everything else should come out after optimization as desired for perfect coupler. For rat race, one line should be 3 times larger then other 3.

My only concerne isthat simple, one section branch coupler could not be good in such a wide frequency band, so you might have to apply multisection design.

flyhigh

Re: branchline coupler

hi,
your answer helped me alot. i´m working with serenade 8.5 sv.
now i reach more then 90° to be exact always around 200°. is there a possiblity to reach the 90° at 1, 1.5, 2 and 2.5 GHz exactly.

now i put for s11 to s44 -60 dB. i was thinking of to say s31 and s41 are -3dB but that didn´t help me either.

thanks for help

Re: branchline coupler

Hi,

sorry bibijons, I didn'i uderstand if you have succeed in optimizing coupler or not? 200deg is a big discrepancy, it might indicate that your lines dimensions are not all equal and you got rat race instead of branch coupler.

Unfortunately I don't have any experience with multisection couplers, but think that this could be the solution for widening the band of operation.

I don't think that can achieve exact phase shift of 90 deg at multiple frequencies. It is almost allways like 90 on one frequency only and then less then 90deg for lower (or higher) frequency and more than 90deg for higher (lower). The correct specification is +/- delta phase in a bandwidth. Anyway, bandwidth for the phase of 90 deg is not that bad, the limiting factor to determine the operating bandwidth of a hybrid is almost always the isolation between the ports (S12, S34 in your case). Reasonable values of isolation are regarded to be -25 to -30dB. Isolation bandwidth is 10-15% for single coupler secton.

flyhigh

Re: branchline coupler

you are right in this frequency range i can not get at each point 90°, so i will design two branchline couplers. i´m not sure how to put these together.

i will see...

maybe i just get one in a store. i heard ANAREn is pretty good.

thx for help

bibijons